The dexamethasone suppression test (DST) was studied in 20 healthy mal
e volunteers, of whom 15 were reinvestigated after 5 to 6 weeks. The m
ean pre-dexamethasone cortisol levels did not differ between the two o
ccasions. Furthermore, they were not correlated. The numbers of abnorm
al DSTs were reduced from 35% to 7%. The poor test/re-test reliability
indicates that unspecified arousal may be a confounding factor in the
outcome of DST.